Review home page | General Overview | Photoshop Elements | Games | Rob's Benchmarks | Other Benchmarks | Conclusion

Review: Dual 2.0Ghz Power Mac G5

small logo

The Short Strokes...

If you don't have a lot of time to read detailed reviews such as this one, all you need to know is this: The Dual 2.0Ghz G5 is a winner. In addition to its compelling design, this machine has the horsepower to handle anything you can throw at it. Whether you're looking to push pixels, make movies, or gorge out on games, the Dual G5 will get the job done.

My advice is, if you need the speed now and you have the money now, go buy one. If you need even more speed, wait a year or so for the dual 3.0ghz machines. And if you have the need for speed but lack the cash, consider mortgaging your home; selling unused assets like your TV, stereo, and car; or taking a night job to raise the necessary cash. This machine is like no Mac I've ever used. As with all advice, there are caveats, of course. The G5 will likely have a few compatibility issues, and its new PCI slot architecture may render all of your current PCI cards worthless. But the G5 still represents a tremendous advance in the state of Mac computing, and is worth strong consideration if you have the need for speed...

[Note: the above desktop image is one of many from Rob Randtoul's amazing galleries at PlasmaDesign ... the machine and monitor are borrowed from Apple's G5 Graphics page.]

The Machines

For all comparison tests in this report, I used the new Dual G5 against my 2001 edition G4/733. Each machine's basic specs are as follows:

  • G4/733: OS X 10.2.8, 1.5gb RAM, 40gb original IDE hard drive, NVidia GeForce3 video card (64mb VRAM)
  • G5/2.0Ghz: OS X 10.2.8, 1.5gb RAM, 160gb serial ATA hard drive, ATI 9800 Pro video card (128mb VRAM)

All tests were set up with no "Sharing" services (web, FTP, etc.) active, and no additional applications were running during any tests. This doesn't mean my results are scientific by any stretch (I hand-timed most of them, leading to errors on the order of a half-second or so either way), but they are consistent.

The Details

I looked at the G5's performance from a number of fronts: in general, how does it feel to use it day to day (Section 1), how does it do with Photoshop Elements (Section 2), how is it as a games machine (Section 3), and then crunching through two different sets of benchmarks (Sections 4 and 5). Click any section title below to jump to that page of the report.

Please note that some of these pages are somewhat large. I used Keynote to create all the charts and graphs, and didn't spend a ton of time on making them as small as possible. If you're on a broadband connection you probably won't notice; if you're on a modem, you probably will. If you'd rather just see the charts on their own and skip my commentary and other images, I've created a benchmarks PDF [236kb download] containing every chart in this report -- at higher resolution and with more quality than the online versions.

1. General Overview A non-scientific summary of my experience to date with the dual G5. Includes a section on the "out of the box" experience (with photos), along with some general commentary on the machine's design, noise levels (including sound clips), and things to be aware of if you're considering purchasing a Dual G5.

2. Focus: Photoshop Elements Though I'm far from a graphic designer, I do a fair bit of work editing photos and playing around with images. For that, I chose the $99 "second cousin" to Photoshop, Photoshop Elements. Though not accelerated for the G5 like its relative, the results of my homebrewed benchmark test are nonetheless quite impressive.

3. Focus: Games! Play them or not, games drive both hardware sales and new technology. New games require better video cards and CPUs, and these new technologies then further allow game makers to push the performance envelope. So how does the Dual G5 stack up as a gaming platform? Find out in this hands-on report covering a number of titles, including Jedi Knight II, Nascar Racing 2003, Tiger Woods 2003, Tony Hawk Pro Skater 4, the X-Plane flight simulator, and (of course!) the mandatory Quake3 benchmarks.

4. Rob's Unofficial Benchmarks When 10.0 came out, I created a benchmark page that looked at application launch times, Perl's speed, and a number of routine day to day tasks that I perform with my machine. I've now updated the chart to reflect 10.2 on both the original G4 test box as well as the new Dual G5. Not surprisingly, the Dual G5 creates some fairly eye-opening results. Panther results will be added as soon as the final OS arrives at my house, hopefully on the 24th of October.

5. Standardized Benchmarks In addition to my homebrew benchmarks in section four, I also ran a few widely available test suites on both machines, including XBench, SpeedRun, Cinebench 2003, and the aptly named Let1KWindowsBloom. I don't believe any of these have been adjusted (if necessary) for the G5 as of yet, but the results are still insightful.

6. Conclusion After all the testing and hands-on time I've had in the last two weeks or so, I think I've come to know this machine fairly well. Read the conclusion to see what I think Apple did right, what they did wrong, and what this box might mean for the future of the Mac...

Review home page | General Overview | Photoshop Elements
Games | Rob's Benchmarks | Other Benchmarks | Conclusion

[Feedback is welcome -- please use this thread on our forum site to discuss this article.]

Counter from

Contact Us | ©2003 ... reproduction allowed with full attribution (including hyperlink) to source.